Evidence for a KATP Channel in Rough Endoplasmic Reticulum (rerKATP Channel) of Rat Hepatocytes

We report in a previous study the presence of a large conductance K+ channel in the membrane of rough endoplasmic reticulum (RER) from rat hepatocytes incorporated into lipid bilayers. Channel activity in this case was found to decrease in presence of ATP 100 µM on the cytoplasmic side and was totally inhibited at ATP concentrations greater than 0.25 mM. Although such features would be compatible with the presence of a KATP channel in the RER, recent data obtained from a brain mitochondrial inner membrane preparation have provided evidence for a Maxi-K channel which could also be blocked by ATP within the mM concentration range. A series of channel incorporation experiments was thus undertaken to determine if the ATP-sensitive channel originally observed in the RER corresponds to KATP channel. Our results indicate that the gating and permeation properties of this channel are unaffected by the addition of 800 nM charybdotoxin and 1 µM iberiotoxin, but appeared sensitive to 10 mM TEA and 2.5 mM ATP. Furthermore, adding 100 µM glibenclamide at positive potentials and 400 µM tolbutamide at negative or positive voltages caused a strong inhibition of channel activity. Finally Western blot analyses provided evidence for Kir6.2, SUR1 and/or SUR2B, and SUR2A expression in our RER fractions. It was concluded on the basis of these observations that the channel previously characterized in RER membranes corresponds to KATP, suggesting that opening of this channel may enhance Ca2+ releases, alter the dynamics of the Ca2+ transient and prevent accumulation of Ca2+ in the ER during Ca2+ overload.


Introduction
Ion channels are present in endomembranes. These channels are thought to play an important role in cellular processes such as compensation for electrical charges, generation of a pH gradient [1], oxidative stress production [2,3], and cell volume regulation.
Potassium channels have also been found in endo/sarcoplasmic reticulum membranes. Evidence for a K + selective channel of high conductance was provided by Picard et al. (2002) [4] using sarcoplasmic reticulum (SR) membrane preparations from human (193 pS) and sheep (185 pS) atrial cells. The gating behaviour and permeation properties of this channel were unaffected by the addition of 4-AP and iberiotoxin (IbTx; a specific blocker of Maxi-K Ca channels) [4]. Voltage-gated potassium channels were also identified in the SR of diaphragm [5] and frog skeletal muscle [6]. In diaphragm SR the channel conductance and gating properties were not affected by physiological concentrations of Ca 2+ , Mg 2+ , and the channel appeared insensitive to glyburide (a selective blocker of ATP-sensitive K+ channels), and charybdotoxin (ChTx) (blocker of Ca 2+ -activated K + channels) [5]. It was proposed on the basis of these observations that SR K + channels could serve in maintaining the Ca 2+ homeostasis [5,7].
In contrast to the SR, where voltage gated potassium channels have clearly been identified, there is little evidence for voltage activated K + selective channels in the rough endoplasmic reticulum (RER) of non-excitable cells such as hepatocytes. We have shown in a previous work that the RER membrane of hepatocytes contains a large conductance (509 pS) K + channel activated by voltage and inhibited by 4-AP (5 mM) [8]. Notably channel activity was decreased in presence of ATP 100 μM on the cytoplasmic side and totally inhibited at ATP concentrations greater than 0.25 mM [9]. These observations argue for the potential presence of a K ATP channel in the RER membranes of hepatocytes. It was reported however that the activity of several Maxi-K Ca channels could be altered by intracellular ATP [10]. Of particular interest is the observation that the Maxi-K Ca channel we identified in a mitochondrial inner membrane preparation from rat brain was inhibited by mM ATP [11]. These results thus raise the possibility for the presence in internal organelles, including RER, of large conductance K + channels sensitive to ATP not corresponding to K ATP type channels. A study was thus undertaken to determine the nature of the ATP sensitive channel we identified in RER of hepatocytes. Our results indicated that the large conductance channels we characterized in RER of hepatocytes correspond to a K ATP channel.

Ethical statement
All experiments were executed in accordance with the Guide for Care and Use of Laboratory Animals (National Institute of Health Publication No.80-23, revised 1996), as approved by the Research and Ethics Committee of Shahid Beheshti University of Medical Sciences (1100/1-87/ 11/21; 2008). Male Wister rats (180-210 g) were housed in a controlled environment (temperature 22 ± 2 _C, humidity 50 ± 10% and 12:12-h light/dark cycle) and were acclimatized for a week before use in experiments. Rats had free access to food and water ad libitum.

Preparation of membrane vesicles
Rough microsomes derived from RER of rat liver cells were prepared as previously described [13,14]. Briefly, three rats were anesthetised and euthanized by decapitation and then, livers were rapidly removed and homogenized in 30 ml of an ice-cold sucrose (0.25 M) solution at 2850 RPM using potter homogenizer (stage 1 for Western blotting). After adding 60 ml of icecold sucrose (0.25 M) solution, the homogenate was centrifuged at 8700 ×g for 13 min (stage 2 for Western blotting). Thereafter, the supernatant was centrifuged at 110000 ×g for 60 min at 4°C (Beckman model J-21B, USA). After dissolving the pellet in 9 ml of ice-cold sucrose 2 M, the solution was transferred to a 30 ml glass homogenizer, and was manually homogenized 8-10 times to obtain a homogenous suspension. The suspension was subsequently centrifuged at 300000 ×g for 60 min in a sucrose gradient, and the resulting pellet dissolved in 20 ml of sucrose 0.25 mM + imidazole 3 mM + Na pyrophosphate 0.5 mM (stage 3 for Western blotting), and centrifuged three times at 140000 ×g for 40 min. The resulting pellet (RER microsomes) was dissolved in 1 ml sucrose 0.25 mM + imidazole 3 mM at a final concentration of 7 mg/ml (stage 4 for Western blotting). Rough microsomes were stored in 10 μl aliquots in 250 mM sucrose/3 mM imidazole, pH 7.4 at -80°C for one month.

Bilayer formation and vesicle fusion
Planar phospholipid bilayers were formed in a 300 μm-diameter hole. The cis (cytoplasmic side) and trans (luminal side) chambers held 4 ml of KCl 200 and 50 mM (pH 7.4), respectively. Planar phospholipid bilayers were painted using a suspension of L-a-lecithin in decane (25 mg/ml). Fusion of the vesicles was initiated mechanically by gently touching the bilayer from the cis side.

Recording instrumentation and statistical analysis
Single channel currents were measured with a BC-525D amplifier (Warner Instrument, USA). The cis chamber was voltage-clamped relative to the trans chamber, which was grounded. Electrical connections were made by using Ag/AgCl electrodes and agar salt bridges. All recordings were filtered at 1 kHz digitized at a sampling rate of 10 kHz and stored on a personal computer for off line analysis using PClamp9 (Axon Instruments Inc, USA). Unitary channel conductance was calculated from the current-voltage relationship. Open channel probability (P o ) was calculated using the standard event detection algorithms in PClamp9. P o was calculated from segments of continuous recordings lasting 50 s. The significance of differences was determined by Student's t test. Data are expressed as mean ± S.E. (standard error).

Results
The nature of the large conductance K + channel we previously identified in the RER membrane of hepatocyte [8] was investigated in a series of experiments aimed first to characterize the channel pharmacological profile, and second to establish if the K ATP Kir6.x and SUR subunits were present in our RER membrane preparation.
Effect of a nonspecific K + channel blocker, TEA +

Effect of calcium ions and Ca 2+ -dependent K + channels blockers, ChTx and IbTx
We have already shown that the large conductance K + channels in RER are inhibited by ATP in a dose-dependent manner [9]. Since, both Maxi-K Ca and K ATP channels are sensitive to ATP, we considered the possible effect of Ca 2+ ions and Maxi-K Ca blockers on the K + channel from RER membranes. Fig 2A-2C show examples of single channel recordings obtained at -20 mV. Changes in channel activity were not observed after addition of 1 mM EGTA, arguing for the K + channel we identified in the RER not being Ca 2+ sensitive (Fig 2A). In addition, cis addition of IbTx (1 μM; n = 5), a specific inhibitor of Maxi-K Ca channels, and ChTx (800 nM; n = 4) failed to modify the channel open probability and amplitude (Fig 2B and 2C). These results are summarized in the bar graph presented in the right panel. Altogether these observations provide evidence for K + channel in RER not corresponding to a Max-K Ca channel.

Effect of sulfonylureas on channel activity
In additional experiments, the effect of glibenclamide and tolbutamide, two well-known sulfonylureas K ATP channel blockers, was examined on RER channel activity. Fig 3A shows that the addition of 10 and 25 μM glibenclamide on the cis side had no significant effect on the channel Po and current jump amplitude at 0 mV (n = 2), while glibenclamide at 50 μM caused a small but significant decrease in the amplitude of the current jump and channel P o . In contrast a total block of cannel activity was observed with 100 μM glibenclamide in the cis chamber at positive but not negative potentials (Fig 3B). These results are summarized in the bar graph presented in the lower panel.
Complementary to these observations the results presented in Fig 4 show that the K + channel from RER was completely inhibited by addition of tolbutamide (400 μM) to the cis compartment at negative and positive potentials (n = 5). These observations argue for the K + channel in RER having the characteristics of a K ATP channel.

Western blot analysis of K ATP channel subunits
A Western blot analysis was next conducted to determine if K ATP subunits were present in microsomes prepared from rat hepatocytes. We first evaluated the purity of our microsome preparation using antibodies directed against various cellular proteins considered unique to particular subcellular regions ( Fig 5A). As seen in Fig 5B, the homogenate (step 1) contains as expected mitochondrial inner membrane the quantity of which was substantially reduced in final step (step 4) of RER membrane preparation.
We next probe for the presence of K ATP subunits in our RER microsome preparations using antibodies directed against the K ATP Kir6.2, SUR1, SUR2A, and SUR2B subunits. Labelling with the anti Kir6.2 antibody showed a specific band at 55 kDa, a molecular weight corresponding to the molecular weight estimated for Kir6.2 (Fig 6A) (n = 4). The results presented in Fig 6B indicate in addition that the anti-SUR2A antibody could labelled a 150 kDa band in crude and RER fractions, whereas two distinct bands at 100 kDa and 150 kDa were detected using the anti-SUR2B and anti-SUR1 antibodies in both the homogenate and RER membrane preparation respectively. These data confirm the presence of the K ATP Kir6.2, SUR1 and SUR2 A and B subunits in the RER membrane fraction used for channel incorporation. We also tested for the presence of BK Ca channel protein in the endoplasmic reticulum membrane [15,16]. The immunoblot presented in Fig 6C indicates in this regard that the BK Ca channel α subunits protein is expressed in our endoplasmic reticulum membrane preparation, despite the absence of detectable BK Ca channel activity in our incorporation experiments.  Discussion K ATP channels were originally discovered in the heart, and have since been found in numerous tissue preparations [17,18]. K ATP channels are also present in endomembranes, in particular in mitochondria (mitoK ATP ) [19]. The main functional feature of K ATP channels arises from rerK ATP Channels their sensitivity to internal ATP levels, with K ATP being blocked at submilimolar ATP concentrations. Internal block by ATP is not however exclusive to K ATP channels. In fact, several studies have shown that Maxi-K Ca channels could be inhibited by intracellular ATP [10,20,21]. In this regard we have presented evidence that a Maxi-K Ca channel in the inner membrane of mitochondria could be blocked by internal ATP in the mM rage [11]. These observations casted some doubt on the actual nature of a K + channel of large conductance we identified in a RER membrane preparation from hepatocyte as this channel could be inhibited by ATP within the same concentration rage [8,9]. We demonstrated in this work through Western blotting analysis and pharmacological characterization that the K + channel of large conductance identified in RER membranes possesses features associated to a K ATP type channel The K + channel blocking agent TEA + was first used to probe the rerK + channel conduction pathway. TEA acts from both sides of the channels and its blocking action is voltage-dependent [22,23]. Our results showed that 10 mM TEA + decreased unitary current amplitude of the K + channel from RER (RERK + ) at positive voltages, a result compatible with a fast block mechanism. These results do not however enable one to discriminate between K ATP and Maxi-K Ca channels as TEA + was found to block Maxi-K Ca channels [20,21], albeit at lower concentrations than those reported to block K ATP .
Several pharmacological blockers of Maxi-K Ca channels are known, including the scorpionderived peptides ChTx and IbTx [24,25]. In the current study, we showed that addition of IbTx and ChTx to the cis side did not affect the channel activity, thus confirming that the channel does not belong to Maxi-K Ca channel family. This conclusion is further supported by the observation that the channel remained active in zero Ca 2+ conditions. Notably the high conductance Slack ('sequence like a Ca 2 +-activated K + channel; also termed Slo2.2) and Slik (also termed Slo2.1) [26], sodium-activated K + (K Na ) channels [27][28][29] have been documented to be insensitive to the Maxi-K Ca channel blocking agent iberiotoxin but inhibited by TEA + at the millimolar (10 mM) concentration range [26,28]. More importantly, application of ATP to the cytoplasmic face of Slick channels was found to reduce channel activity [28]. These features are in line with the pharmacological profile we obtained for the K + in RER membranes. It has been suggested that the Na + concentration required for channel gating in both cases ranges from 7 mM to 180 mM [29]. As all our experiments were performed at contaminant levels of Na + ( 100 μM), the possibility that the K + channel from RER corresponds to either Slo2.1 or Slo2.2 can be ruled out. However observations by Kim et al. (2007) [30] indicated that K Na channels could not be inhibited in the presence of a K + channel blocker cocktail containing a mixture of Ba 2+ , glibenclamide, 4-AP, apamin, and quinidine [30], a result incompatible with the blocking effect of glibenclamide we observed on the K + channel from RER membrane.
Sulfonylurea compounds are known to bind to the SUR subunit and inhibit the K ATP channel activity [31,32]. To determine if the endoplasmic reticulum potassium channel in rat hepatocytes corresponds to K ATP channel, we tested the effects of a cis application of tolbutamide and glibenclamide on channel activity. Fig 4 demonstrates that the addition of 400 μM of tolbutamide to the cytoplasmic side (cis chamber) completely blocked channel activity. Furthermore, our result showed that glibenclamide at 100 μM inhibited the channel activity at positive potentials, but failed to affect the channel unitary current amplitude and open probability at negative potentials. The mechanism by which glibenclamide inhibits the ATP sensitive K + channel in RER remains to be established. It has been suggested that glibenclamide was not blocking the channel directly but might act by binding either to the voltage gate and/or to the inner mouth of the channel leading to enhanced hydrophobic interactions between the two and a stabilization of the channel inactivated state [33]. Inhibition of channel activity by glibenclamide and tolbutamide was in the present work observed at micromolar concentrations. Notably, the K ATP channel sensitivity to sulfonylurea derivatives was found to vary over a three orders of magnitude depending on the cell type. For instance K ATP channels in pancreatic βcells, cardiac muscle, and brain were found to be inhibited by glibenclamide at nanomolar concentrations, whereas K ATP channels in other cells including epithelial cells appeared more glibenclamide resistant with concentrations for half-inhibition in the micromolar to millimolar ranges [34][35][36]. It has been suggested that the cytoskeleton affected K ATP channel sensitivity to sulfonylureas inhibition, with a disruption of cytoskeleton structure by DNaseI leading to a 20 folds decrease in sulfonylureas affinities to K ATP [37]. Because ER extraction involves disruption of the cell cytoskeleton it is possible that the low affinity we observed for glibenclamide and tolbutamide induced channel inhibition partly arised for the absence of cytoskeleton. Altogether, our pharmacological results would support the presence of a K ATP channel in the hepatocytes RER membrane.
K ATP channels are octomeric structures comprising four pore-forming Kir6.1 or Kir6.2 subunits and four sulfonylurea receptors (SURs) encoded by two distinct genes, SUR1 and SUR2 (SUR2A and SUR2B). Kir6.2 was reported to be expressed in ER fractions of pancreatic beta cells [38] and cardiomyocytes [39]. Our current Western blot results also argue for Kir6.2 being present in the RER membrane of hepatocytes and for the presence of the SUR1, SUR2A, and SUR2B subunits (Fig 6). Our Western blot results are unlikely to be explained by the presence of contaminants coming from other subcellular compartments, as we failed to detect the expression of plasma membrane or Golgi matrix specific proteins in our microsomal preparations. To our knowledge, this is the first study addressing the combination of SURs subunits in endoplasmic reticulum of rat hepatocytes.
There is evidence that the sensitivity of the channel to different sulfonylureas could depend on the SUR isoform. For instance, Kir6.2-SUR1 but not Kir6.2-SUR2A channels are blocked by tolbutamide with high affinity [40,41]. Glibenclamide blocks both types of channels with high affinity. In line with these observations, reconstitution of K ATP channels by co-expression of Kir6.2 and SUR2 in COS cells, and sulfonylurea binding measurements have revealed that despite similar structural features between SUR2 and SUR1, SUR2 binds glibenclamide with a lower affinity compared to SUR1, demonstrating that different responses to sulfonylureas may be linked to the SUR isoform [42]. These data, added to the fact that glibenclamide consists in a tolbutamide moiety plus a non-sulfonylurea meglitinide group known to inhibit K ATP channels [43,44] suggest that SUR1 contains a high-affinity tolbutamide-binding site missing in SUR2A. Our Western blot analysis provided unambiguous evidence for the presence of the Kir6.2 and SUR2A K ATP subunits in the RER membrane of hepatocytes. Our analysis also showed two bands of 150 kDa and 100 kDa for the SUR2B and SUR1 subunits. It has been reported that the antiSUR2B anti-body recognises polypeptides of different molecular weights (150, 100, and 55 kDa) in some microsomal fractions including microsomes from rat heart sarcolemmal membranes [45]. More importantly, since the sequences of antigen peptide in SUR2B are highly homologous to the corresponding region in SUR1, we cannot currently rule out the possibility that the bands assigned to SUR2B can also be assigned to SUR1 [45]. Our results thus suggest that Kir6.2 and SUR2A are expressed in RER membrane of hepatocyte together with SUR2B and/or SUR1. There are supporting data that SUR1 and SUR2A coassemble readily and randomly to form heteromeric K ATP channels [46]. Furthermore, Wheeler et al. demonstrated that SUR1, SUR2A, and SUR2B can coassemble in all the possible pair-wise combinations to form functional K ATP channels with distinct properties, resulting in an enhanced functional and pharmacological diversity [46]. Determination of the Kir6.2 and SUR subunits exact combination will require further investigation.
Several groups [15,16] including ours (present work) have provided evidence for the presence of the BK Ca channel α-subunit (pore-forming subunit) in endoplasmic reticulum membrane. Notably channel activity could not be detected at single channel level in these preparations. These observations contrast with results obtained from mitochondria and nucleus membranes where BK Ca channel activity has been documented [47,48,11]. We cannot currently rule out the possibility of a non functional BK Ca isofrom present in our RER membrane preparation. Confirmation of this proposal will await further work.
The precise functional role of K ATP channels in RER is currently unknown. Both SR/ER Cland K + channels act as counter transport systems during rapid Ca 2+ release and uptake as to keep electroneutrality and maintain the electrochemical driving force acting on Ca 2+ ions [49,50]. Given that Ca 2+ regulations by the ER is prominent in cellular apoptosis [51,52], these data suggest that K ATP in RER membrane channel may contribute to the regulation of ERmediated cellular mortality so that a defect in this regulatory process may cause a progressive loss of cell function and the generation of a cellular pathological state.